Megakaryocyte/Platelet-Restricted FVIII Expression Using a Platelet Factor 4 Promoter: Determination of Activation State and In Vivo Efficacy during Thrombocytosis.

Ectopic delivery of coagulation factor VIII (FVIII) to megakaryocytes (Mk) represents a viable approach for localized tenase generation by effectively concentrating the FVIIIa/FIXa enzyme-cofactor complex onto the negatively-charged phospholipid surface of activated platelets. While phenotypic correction has been demonstrated using hemophilia A (FVIII^−/−) murine models in vivo, the activation state of platelet FVIII (pFVIII), optimal promoter choice, and phenotypic correction in the setting of a thrombocytotic stimulus remain unestablished. Preliminary microarray experiments using human platelets (N=5) demonstrated that the Mk-specific platelet factor 4 (PF4) transcripts were among the most abundant, prompting use of the 1.1 Kb PF4 promoter for Mk/platelet-restricted expression of human B-domain-deleted (hBDD) FVIII within the background of an exon 17-deleted mouse model of hemophilia A (PF4/hBDD/FVIII^−/−). A chromogenic tenase assay using gel-filtered platelets from PF4/hBDD/FVIII^−/− mice confirmed the presence of functional FVIII equivalent to 73 mU·1x10⁹ platelets·mL⁻¹ (N = 10 mice). In contrast, FVIII was not detectable in PF4/hBDD/FVIII^−/− plasma (assay sensitivity <20 pM) or in platelets from FVIII^−/− mice. The ectopic pFVIII did not affect the release and/or function of other a-granule storage proteins as established by parallel measurements of platelet factor V (FV) using a prothrombinase assay. Paired tenase assays (± thrombin) confirmed that pFVIII (unlike pFV) required thrombin cleavage for complete activation. To delineate the effects of a thrombocytotic stimulus on pFVIII expression and/or function, PF4/hBDD/FVIII^−/− (N = 10) or FVIII^−/− control mice (N = 5) were injected with thrombopoietin (TPO; 10μg/kg/day for 5 days) resulting in an 87% average increase in platelet count. Day 10 tenase assay of TPO-injected PF4/hBDD/FVIII^−/− mice demonstrated a 66% reduction in pFVIII activity (25 mU FVIII·1x10⁹ platelets·mL⁻¹), unassociated with altered expression of a-granule-stored amyloid-b-precursor protein (AbPP) as a control for storage granule content; plasmatic FVIII remained undetectable. In contrast, the decrease in total platelet FVIII biomass was less pronounced in TPO-treated PF4/hBDD/FVIII^−/− mice, representing a 35% reduction from 147.6 mU to 96 mU after TPO stimulation. The decreased pFVIII in TPO-stimulated PF4/hBDD/FVIII^−/− mice correlated with loss of phenotypic correction as evaluated using tail bleeding survival rates: wild-type mice (100%; N= 5), FVIII^−/− (0%; N=5), PF4/hBDD/FVIII^−/− (TPO-naïve 60%; N=11), PF4/hBDD/FVIII^−/− (TPO-treated 0%; N=7) (p value between PF4/hBDD/FVIII^−/− mice with and without TPO stimulation = 0.002). While these data establish that Mk-directed pFVIII (unlike pFV) is proteolytically inactive upon platelet activation, they also imply that thrombocytotic stimuli negatively affect the pFVIII bioavailability and phenotypic efficacy. As importantly, the hemostatic efficacy of platelet FVIII correlates best with localized platelet FVIII delivery (FVIII concentration/platelet) and not the systemic platelet FVIII bioavailability.